There is an article published in the BMJ (allstat does'nt accept attachments, sorry) on left handedness and risk of breast cancer.
It recalls an article that was published a number of years ago in the NEJM claiming that left-handed people had a shorter life expectancy.
The result in the breast cancer study is valid only for pre-menopausal women with normal BMI.
I don't claim to be an expert, but might this be a case of analyzing sub-samples until the (0.05) jackpot is hit?
If enough tests on a large cohort are run, something is bound to turn up.